Moore reconstruction and MuonID

1

and the HLT
Gabriella Cataldi (INFN Lecce) Michela Biglietti
(Universita di Napoli-Federico II)
2
Outline
3
Standard Moore flow

• Each step is driven by an Athena top-algorithm
• Transient objects are passed via TDS/StoreGate
• Independent algorithms, the only coupling is
  through the transient objects

MooAlgs
RPC/TGC digits
MooMakePhiSegments
PhiSegments
MooMakeRZSegments
MDT digits
MooMakeRoads
CrudeRZSegments
MooRoads
MooMakeiPatTracks
MooiPatTracks
MooStatistics
MooMakeNtuples
Ntuples
4
Reconstruction Strategy

• Searches for ? regions of activity
• From the RPC/TGC ? measurements ?- Segments
  are created
• Searches for R-Z regions of activity
• For each ?-Segment, the associated MDTs is
  found
  and a crude RZ Segments
  
  is built
  
  (essentially collections of z hits) .

5
Reconstruction Strategy

• Pattern recognition and outer Roads
• Inside MDTs the drift distance is calculated from
  the drift time, by applying various corrections
  on it (TOF, second coordinate, propagation along
  the wire, Lorenz effect). From the 4 tangential
  lines the best one is found.
• All the MDT segments of the O station are
  combined with those of the M layer. The MDT hits
  of each combination are added to the phi-hits of
  the Phi Segment, forming outer track
  candidates. All the successfully fitted
  candidates are kept for further processing.

MDT mutilayer

• Final tracks
• associate inner station MDT hits. A final track
  is defined
• The successful outer track is subsequently
  used to
• as a successfully fitted collection of trigger
  hits and of MDT hits from at least two layers.

6
Moore is a container
Athena algorithms with different features/goals
Moore structure
MooLVL2Algs
MooAlgs

• MooAlgs contain the standard
• sequence of Algorithms
• MooMakePhiSegments
• MooMakeRZSegments
• MooMakeRoads
• MooMakeTracks
• MooLVL2Algs contain a possible
• seeding approach
• MooMakePhiSegmentLVL2
• MooMakeRZSegmentLVL2

MooStatistics
MooAlgs_n
MooEvent
Events for reconstruction
7
MooLVL2Algs strategy
run atrig
trigout10GeV.txt trigout20GeV.txt .. .. ..
only on barrel
ASCI files
Set-up the logic link TestRelease/
/run/Persint.out ? / /trigoutxxx.txt
run athena with the appropriate jobOptions.txt
RPC digits
MooMakePhiSegmentLVL2
PhiSegment
MooMakeRZSegmentLVL2
MDT digits
RZSegmentMap
How to run Moore and MooLVL2Algs Documentation
in http//people.na.infn.it/bigliett/moore/Moore
Doc/MooreDoc/
8

• MooLVL2Algs
• outputs are
• PhiSegments
• RZSegments

• In order to perform a
• fit the standard Algorithms
• MooAlgs must be plugged
• in the sequence

MooMakePhiSegmentLVL2MooMakeRZSegmentLVL2MooMake
RoadsMooMakeTracks
9
First need

• If we want to have the seeded Moore in the HLT
  environment we need
• MooAlgs
• MooLVL2Algs

First step
10
How to seed MooLVL2Algs.

• As it is now
• MooLVL2Algs are seeded with the list of
  HitsOnTrack from LVL2.
• MooLVL2Algs work with Old EDM layout M geometry.
• In the future (personal opinion)
• MooLVL2Algs for HLT should use the track
  information, opening a Region of Interest around
  the track, and performing the PhiSegment,
  RZSegment search, as for standard Moore, only in
  a limited region.
• The new ID scheme and the new EDM help in
  these functionalities.
• MooLVL2Algs functionality with list of
  HitsOnTrack will be useful for comparison of
  results

RPC hits
MDT hits
11
Hits comparison

• Same dataset used for MooAlgs MooLVL2Algs
• The 2 different sets of algos are chained
  together
• Fit results from the same sample of events are
  directly compared

Plans for testing a more general hit addition
procedure
12
c2
13
1/pT pull
14
pT resolution
15
Plans for testing a more general hit addition
procedure

• Hit Addition in Moore starts from RZSegments.
• The RZSegments in MooLVL2Algs are built from the
  output of LVL2 Trigger not from the collection
  of digits.

.thinking over
16
Moore (for HLT) Vs ATLAS releases

• Moore is working in version 5.3.0 with OLD EDM,
  new ID scheme, new Geometry (able to read TDR and
  DC1 data)
• Moore working version with new EDM has been
  committed, but not yet in release. (Ready for tag
  in 6.0.0)
• With the new EDM release Moore has been fully
  restructured

17
Commit in 5.2.0

• Bug fix (see Stefanias talk)
• Adoption of DataVector instead of handle at
  EndOfEvent
• No more skip pick parameters in
  jobOptions.txt.
• Use EventSelector.EventNumbers instead
• jobOptions Moore_jobOptions.txt

18
New EDM Moore fully restructured.
Giorgos is writing and redesigning a large part
of the code
A crazy(?) idea in order
MooLVL2Algs
MooAlgs

• to keep the
• functionality of the
• code as it is now
• to have the possibility
• to compare the
• results and the
• performances (chain)

MooStatistics
MooNefeliAlgs
MooEvent
MooNefeliEvent
19
(No Transcript)